Table of contents

Nanometre-sized particles of ferrite, commonly used in magnetic fluids, are single-domain. The direction of magnetic moment of these small, uniaxial, ferromagnetic particles is known to fluctuate due to thermal agitation, and can relax through the Neel-type relaxation mechanism. The relaxation time of such fluctuations is usually determined by means of Brown's equations for high and low barrier heights. More recently, modified equations catering for a continuous range of barrier heights have been proposed. Comparison of these equations shows that, even in the most extreme case only a factor of approximately 1.7 distinguishes the corresponding eigenvalues (which represent the inverse of the relaxation time). It is concluded that the major source of error in predicting the relaxation time arises, not primarily due to the particular equations used, but because of the large uncertainty in obtaining precise experimental data needed to determine the components, f₀ and sigma , of these equations. For example, for a small change in anisotropy constant K by a factor of 2.5 (typical values for the system considered here are (2-5)*10⁴ J m^-3), the calculated values of Neel relaxation times using Brown's equation differ by a factor of about 37, corresponding to times of 1.6*10^-7 to 4.3*10^-9. An experimental value of 5*10^-9 s determined from the frequency of the maximum of the loss-peak of the imaginary part of the complex susceptibility is at the outer limit of these calculated values.

Magnetic viscosity measurements consist of placing an ensemble of small magnetic particles in a very large magnetic field, which tends to align the magnetic moments, m_p, of the particles in the direction of the field and then observing the decay, with time, of the magnetization when the field is removed. In measuring the magnetic viscosity, it is possible for magnetization data to be unavoidably lost due to the time delay between removing the applied field and making the first magnetization measurement. It is shown that, by taking the inverse Fourier transform of the frequency-dependent complex susceptibility, chi ( omega )= chi '( omega )-i chi ''( omega ), of the sample, it is possible to recover data corresponding to that which would have been measured during the period of the time delay.

A technique that enables one to investigate the magnetization decay of a ferrofluid by means of processed complex susceptibility data is presented. The presence of a particle size distribution, and hence a distribution of relaxation times, is accounted for by means of the Frohlich distribution function.

The scattering of electromagnetic waves by small particles distributed over an interface between two media with different complex dielectric constants is investigated in a dipole approximation. Analytical expressions for the scattered field for different polarization states of the incident radiation are derived. These are shown to reduce in the appropriate limit to those predicted by first-order Rayleigh-Rice theory for scattering by a slightly rough surface. Results for particles on a tilted interface are also derived and used to model scattering by a roughened undulating surface. Formulae are evaluated for a variety of particle shapes and media dielectric constants, and the effect of clustering is briefly examined.

An analysis of the parametric instability which forms the basis for operation of a ubitron-type free electron laser with a longitudinal electric wiggler is presented. Excitation of a space charge wave and a predominantly transverse magnetic waveguide mode results from passage of a relativistic electron beam through a static spatially periodic longitudinal electric field. The beam is contained in a cylindrical metal waveguide and an axial magnetic field of arbitrary magnitude. Effects of the magnetic field on the growth rate, efficiency, radiation wavelength and the required pump wavelength are studied. Suitable values of the magnitude of the magnetic field and other parameters for generating millimetre wavelength radiation are indicated.

A new, segmented-electrode, high-voltage, hollow-cathode, Cu II laser, operating on the 780.8 and 782.5 nm transitions, is studied. Parametric measurements are presented for the discharge voltage, the spatially averaged copper-atom concentration and, for the two transitions, the small-signal gain coefficient, the laser threshold current and the laser output power. The dependence of the laser output power on the active length, the buffer-gas pressure and the transmission of the output coupler is modelled using equations for a low-loss, homogeneously broadened, standing-wave laser. For the 780.8 nm line, the lowest threshold current is 0.19 A for a 5 cm active length and a 0.1% transmission output coupler, and the maximum output power is 58 mW for a 3 A discharge current, a 7.5 cm active length and a 2% transmission output coupler. This performance exceeds that of conventional hollow-cathode lasers on the same transition, the improvement being attributed to the increased operating voltage, which gives a more efficient pumping discharge by simultaneously raising the ground-state copper-atom concentration (through the increased sputtering yield of ions bombarding the cathode) and increasing the efficiency of ionization.

Detection of infrared radiation (10 mu m) by a nonlinear up-conversion technique under a tangential phase-matched situation in AgGaS₂ crystal with Nd:YAG laser as a pump has been demonstrated. A wide angular tolerance has also been obtained with a view to achieving mechanical stability. Its capability of detecting a low-level infrared signal at room temperature and its temperature-dependency have also been demonstrated.

Variations in refractive indices with temperature, the thermo-optic coefficients and temperature-dependent absorption in AgGaSe₂ crystal have been measured. Good agreement between the theory and experiment has been found for the temperature-dependence of phase-matching angles for second-harmonic generation of CO₂ laser radiation in this crystal. Hence variations of the phase-matching angle with temperature in some important nonlinear devices have also been estimated.

Local enhancement of refractive index in potassium-ion-exchanged optical waveguides in glass is of interest for realization of low-loss bends and intersections and more complex structures such as gratings for reflection and phase-matching. Local enhancement may be realized through subsequent ion exchange with an ion having higher polarizability, such as silver. However, in order to design such structures, the diffusion of silver ions into potassium ion-exchanged waveguides must be characterized. In this paper, we present and compare diffusion profiles and coefficients of silver ions in soda-lime glass and in potassium-ion-exchanged waveguides in this glass.

A narrow-bandwidth interference filter has been developed using volume holography in photorefractive LiNbO₃ crystals. The device works as a Bragg reflector using a volume hologram recorded, fixed and developed in a LiNbO₃ sample. With a 2 mm thick LiNbO₃ plate the following performances have been achieved: a peak reflectivity above 32%, a full-width at half-maximum bandwidth below 0.05 nm, an angular field of view of about 3 degrees and a peak wavelength of 518.45 nm (for green Ar laser writing wavelength). These performances compare well with that exhibited by the conventional Lyot-Ohman and Solc filters. A complete explanation of the fabrication procedure is given, including practical details needed for producing such an efficiency with fixed holograms in LiNbO₃.

Stark broadening parameters of Al III lines of interest in the diagnostics of an electrodynamic macroparticle accelerator (rail gun) arc plasma created by evaporation of an Al foil have been calculated using the semi-classical perturbation formalism. Stark widths of Cu IV lines of interest for an arc plasma created by Cu foil evaporation have also been calculated by using the modified semi-empirical method.

The temperature map of a laser-sustained argon plasma is calculated in the gas flow field upstream of a sonic nozzle orifice. The discretized equation of energy conservation is solved with a finite difference method. A forced parallel gas flow field is introduced in order to simulate the gas flow through the nozzle. Such a parallel flow does not account for the pre-cooling effect at the nozzle throat, so that the isentropic laws must be applied at the end of each iteration (outlet) for re-introducing the corresponding gas flux into the inlet conditions of the next iteration. In this way, the gas flux calculation is self-consistent and convergence is obtained after about 200 iterations. The results are compared with experimental data. The agreement is rather satisfactory except that the theoretical laser power threshold is about one order of magnitude higher than the experimental measurement. This is due to the parallel gas flow scheme, which overestimates the gas flux in the plasma core. This effect can be corrected by taking a gas flux value to account for the convergence of the stream tubes toward the nozzle.

Metallic electrodes are used for current conduction in potassium-seeded thermal plasma. Investigations are made on composite electrode surfaces formed due to the adsorption of seed together with the combustion product species. In the metallic copper or stainless steel electrodes used for the experiments, current conduction is restricted to the diffusive mode of conduction. The electrode surfaces are investigated by means of scanning electron microscopic and Auger electron spectroscopic techniques. Differential thermal analysis of the material adsorbed on the electrodes indicates that potassium carbonate is the main constituent of the adsorbate.

This paper presents an approximate formula for the determination of local axis temperatures of an inhomogeneous, axisymmetric, optically thin, local thermodynamic equilibrium plasma column without using the Abel inversion technique. The proposed method is straightforward and is based on recording the spatially integrated radiances of spectral lines. The formula is useful for high gradients thermal plasma columns of the type found in DC electric arcs or plasma jets of gaussian shaped temperature profile. The formulation permits a simple experimental arrangement for rapid monitoring or control of parameters in industrial plasma devices or for the determination of emissivities and line transition probabilities. We also show that the method can be applied to spatially unstable arc discharges, where traditional techniques are difficult and cumbersome to apply.

The formation of structure in the cathode region of non-self-sustained discharge in nitrogen was experimentally studied by means of an image-converter intensifier framing camera, Michelson and Mach-Zender interferometers and a laser schlieren method. The appearance and evolution of the mushroom-like glow structures, which have not been observed earlier in numerical discharge models, were seen. A phenomenological approach was used in modelling the observed structures and the main processes responsible for these formations were revealed on the basis of the density distribution restored from the interferograms.

Experimental measurements of the second derivative d²i/dv² of the probe current i versus probe voltage v were performed in an axially symmetric magnetron sputtering gas discharge plasma in order to determine the electron energy distribution function. The measurements were performed using a cylindrical Langmuir probe oriented in two directions, along and across the discharge axis. The experiments were carried out at a constant argon gas pressure of 3 mTorr and at a constant discharge current of 1 A to a molybdenum cathode. Three different magnetic field configurations in the discharge region were investigated and these were achieved by superimposing an external magnetic field on the original magnetron magnetic field. By varying the magnetic field distribution and, in particular, the respective axial component of the magnetic induction B_z at the probe position, the second derivative d²i/dv² was significantly altered. Certain differences in d²i/dv² were also established with the probe oriented in radial and axial directions. A spherical spatial plasma isotropy of the electron energy distribution function was thus found for the opposed superimposed magnetic field (B_z=-31 G) while for both the original (B_z=36 G) and the supporting (B_z=103 G) magnetic field configurations, a plasma anisotropy was established. The degree of anisotropy was quantified by comparison of the respective average electron energies (E) evaluated for each of the three magnetic field configurations. It was found that assumption of spherical isotropy during determination of the electron energy distribution function leads to overestimation of (E) by up to 10% under the experimental conditions used.

The dependencies of electrical-breakdown time delays t_d on the afterglow period tau were measured and the dependence of the mean values t_d on tau , using the so-called memory curves t_d=f( tau ), was established in nitrogen-filled diodes at 1.3 mbar and 13.3 mbar. The corresponding time-dependent variations of metastable concentration n, i.e. the curves n=F( tau ), were derived, where, at afterglow periods exceeding 1 s, the metastable molecules were predominantly in the N₂(A³ Sigma _u⁺) state. A physical model of metastable de-excitation processes has been proposed and, by solving its differential equations, the following dependencies and constants were obtained: (1) the theoretical dependence of metastable concentration on the afterglow time-the curve n=F( tau ), (2) the range of n₀, the initial concentrations, which are also saturation concentrations of metastables reached in discharges, (3) the metastable concentration-independent constant k in the equation t_d=k/n, (4) the dependence of the half de-excitation time T_1/2 on the afterglow period, T_1/2= phi ( tau ), and (5) the dependence of Z on the afterglow period, Z= Phi ( tau ), where Z is the concentration dependent ratio between the de-excitation rate via binary metastable-metastable collisions and the total rate via all other considered de-excitation mechanisms.

The collective behaviour of weakly ionized RF dusty plasmas is investigated. Negatively charged SiO₂ fine particles are generated and suspended in an SiH₄/O₂/Ar discharge. At high pressure (300 mTorr), the random fluctuations of the discharge and particle motions are suppressed. Coulomb solids with particle diameters from a few micrometres to a millimetre have been observed. At low pressure, the coupling between particles and the background plasma causes large-amplitude low-frequency fluctuations of the dusty plasma.

The kinetics of plasmas sustained by surface wave propagation at a frequency 2375 MHz in argon gas are studied experimentally. A simple manner for estimating the plasma parameters on the basis of general relations is given. The measured electron energy distribution functions are compared to numerical results from a kinetic modelling of surface wave produced plasma. Good agreement is found. By averaging over the measured electron energy distribution functions, the plasma macroscopic parameters (electron density and temperature, as well as frequencies of different processes initiated by electron impact) are obtained.

An experimental study has been carried out into the optical activities associated with surface flash-over on high-voltage insulators in vacuum, aimed at determining the physical origin of the flash-over process. This involves using a 'transparent anode' to form, with combinations of cathode-insulator regimes, a number of electrode configurations, in order to observe the field electron emission and the optical events occurring during surface flash-over. In particular, the flash-over characteristics of two types of insulator shapes, cylindrical insulators and truncated cone insulators, were investigated. The latter allow spatially resolved imaging of the optical activities at two triple junctions and the surface of the insulators. Using this imaging technique in conjunction with the real-time recording function of a video camera system, four types of optical images have been identified. In addition, the influences of temperature and helium conditioning upon the characteristics of a flash-over were studied. Finally, a new model has been proposed to account for these experimental findings, which assumes accumulation of charges at two interface regions at two triple junctions, and that subsequent initiation of flash-over results from the breakdown of interface layers.

Corona simulation by the Popkov/Sarma models involve the Deutsch approximation of the invariance of the electric field direction. The paper implements the models to obtain current density and field profiles at the collector electrode for positive corona in wire-plane, paraxial-cylinder and point-plane gaps in air. The simulations are compared with experimental data and are found to give satisfactory agreement. Consideration is given to reasons why the Deutsch approximation yields low errors.

Electron current stability in relation to filament formation in aluminum plasma at densities from 1 to 1000 atm (10⁵-10⁸ pa) has been studied. Electron momentum relaxation times obtained, including Coulomb and inelastic collisions, were in the picosecond and smaller time scale, and increase in the temperature interval 10⁴-2*10⁴ K. This results in similar electron conductivity temperature dependence. Electron energy losses due to ionization and excitation of atomic particles in the temperature range mentioned above are almost constant. At such conditions electron temperature fluctuations in the plasma must increase, leading to electron current filamentation. The line radiation can work as a stabilizing factor, but it becomes effective only in the nanosecond time scale or at low plasma densities. The process seems to be of importance in relation to the current fragmentation at the arc cathode spot.

This paper is concerned with the theoretical analysis of the compressive squeeze performance of an electrorheological (ER) fluid sandwiched between two parallel circular electrodes, the upper one fixed and the lower one oscillating sinusoidally normal to its own plane. It is shown how the force transmitted through the fluid on the upper electrode varies with amplitude and frequency of oscillation together with applied potential difference and also how the fluid, modelled by a bi-viscous characteristic, responds when electrically stressed compared with its unstressed, Newtonian counterpart. Finally, the relevance of the theoretical results to the determination of the mechanical properties of an ER fluid in squeeze is discussed.

Using the non-local elasticity theory developed by Kroner, Eringen and some others, the interaction between a finite-length crack and a straight screw dislocation parallel to the crack plane is investigated. The following theorem is proved: 'in isotropic, non-local elastic media, the force exerted on a screw dislocation by the other defects is proportional to the gradient of the energy of the dislocation', which is correct in classical elasticity. With the theorem, the non-local image force on a screw dislocation due to a finite-length crack is calculated. The result shows that there is no singularity of the classical image force and there is a maximum of the force at the crack tip.

A modified variational procedure is presented, which allows an efficient determination of dispersion curves and especially of normal modes of phonons in periodic, inhomogeneous materials with abrupt changes of material constants. The Bloch wave eigenvalue problem is solved by a Ritz method using expansion functions with discontinuous first derivatives for the displacements, which satisfy the interface conditions for displacements and stresses. An explicit derivation is given for threefold periodic arrangements of inclusions in a matrix. Questions of convergence are discussed. A numerical comparison with results obtained with customary Fourier trial functions for structures inhomogeneous in one and two dimensions and with exact results for one-dimensional (layer) structures shows the great accuracy of the method.

The polarizabilities and activation energies for surface diffusion of adatoms gallium (Ga), indium (In) and tin (Sn) on a molybdenum (Mo) surface were determined by using a field in microscopy technique. The values of these quantities, i.e. polarizability and activation energy, were obtained by a systematic investigation of the surface diffusion of the adsorbates (adatoms) on a molybdenum tip surface, in an applied electric field and in the absence of an applied electric field respectively. The applied electric field created a driving force on the adatoms, and by measuring the velocity of migration of directional walk of the adatoms and using the theory of directional walk, the polarizabilities were obtained. The activation energies were determined by measuring the displacements of the adatoms on the tip surface, at different values of temperature, using the random walk theory. The values obtained for the polarizabilities and activation energies respectively were 2.71 (AA³) and 0.25 (eV) for Ga, 3.10 (AA³) and 0.30 (eV) for In, and 2.25 (AA³) and 0.50 (eV) for Sn. The values of polarizabilities were less than the values of free atoms.

A new parameter, pressure, has been introduced into study of the electrical properties of polymers, in particular into the thermal breakdown theory for modelling of the breakdown field F_b variations of high-density polyethylene versus pressure. Theoretical F_b values, determined by numerical calculation of the fundamental heat balance equation and the analytical expression of the conductivity, are compared with experimental results obtained in the pressure range 50-500 bar. Good agreement was observed in this comparison.

The paper investigates the impact of the high-field electron diffusivity model in GaAs on the simulated noise performances of metal semiconductor field-effect transistors (MESFETs). On the basis of a comparison between the numerical simulation, carried out through a two-dimensional implementation of the impedance-field method, and the noise power and correlation spectra measured on a 0.6 mu m low-noise MESFETs, it is shown that the intermediate and high-field behaviour of the diffusivity-field curve consistent with the experimental data is close to the one suggested in the study by de Murcia et al. (1991), while other experimental or Monte Carlo data overestimate the diffusivity and therefore the noise.

A design variant of a high-temperature superconducting thin-film switching device controlled by a field-emitted electron beam is proposed. Numerical simulation shows that the superconducting-to-normal state transition time in such a device will be about several hundred picoseconds. Results of analysis of the dynamics of heating and switching of high-temperature superconducting thin film from superconducting to normal state under the action of a surface pulsed energy flow are presented.

Fabry-Perot interferometry is used to study the electro-optical properties of poled poly(methyl methacrylate) and poly(styrene acrylonitrile) polymeric films. The apparent electro-optical effect is shown to be proportional to the piezoelectric coefficients of the film and to the (cos phi ) dipole order parameter. For poly(styrene acrylonitrile), this technique gives piezoelectric coefficients close to those derived from dielectric measurements and thermodynamic considerations. Direct electro-optical activity measurement is used to record the dipole orientation step response function in these two glassy matrices at various temperatures. For both polymers these functions are similar at a given difference in temperature below the glass transition temperature and are non-exponential.

Using Monte Carlo methods, we have calculated depth profiles of X-ray production, Phi ( rho z) curves, for Si K alpha in a Si substrate coated with Cu, Ag and Au films of varying film thickness and electron beam voltages from 10 to 30 kV, as well as the characteristic radiation from films for comparison with August and Wernisch's (1991) analytical approach. Calculation results shows that, for radiation from the film, general agreement is found with August and Wernisch's formula, while it is not found for radiation from the substrate. We demonstrate that the Monte Carlo simulated Phi ( rho z) curves from Si substrate can be well described by the modified Gaussian profile of Packwood and Brown (1981), and the calculated profiles may be parameter-fitted into the analytical formula Phi ( rho z)= Theta ( rho z- rho Delta z)( gamma -(( gamma - Phi (0))exp(- beta rho z)))exp($ - alpha ²( rho z)²).

Bismuth titanate thin films have been grown by the atmospheric pressure metal-organic chemical vapour deposition technique. The influences of substrate temperature and annealing on phase formation and crystalline properties have been studied. (100) oriented Bi₄Ti₃O₁₂ films have been grown at 550 degrees C, while (001) preferred orientation films were obtained at high temperature (about 800 degrees C) annealing. At a temperature of 650-750 degrees C, a film with Bi₄Ti₃O₁₂+Bi₂Ti₂O₇ phases was formed. It is shown that the phase formation may depend upon the growth temperature and composition of film. The orientation of films mainly depended upon the lattice match with substrate.

In order to obtain extremely fine-grained microstructures by crystallization of the metallic glass, the nucleation and growth processes of the crystalline phases have been investigated over a wide temperature range. Analysis of the transformation kinetics and microstructure has allowed determination of the nucleation and crystal growth rates. The parabolic growth rate increases with annealing temperature, and the nucleation rate based on the homogeneous nucleation theory has shown a maximum at an intermediate temperature range. To a certain extent, quenched-in nuclei influence the crystallized microstructures. In this paper, we have also performed an investigation to determine the effect of quenching rate on the grain size and its distribution. The results indicate that the number of crystals per volume unit formed during isothermal annealing depends not only on annealing time and temperature, but also on the quenching rate, that is the number of quenched-in nuclei of the as-quenched amorphous alloy. The results obtained have shown that it is possible to achieve an extremely fine-grained and well-distributed microstructure by controlling the quenching rate and following crystallization of the metallic glass under suitably chosen annealing conditions.

Thoriated tungsten filaments acting as the main part of the cathode of magnetrons used for creation of electromagnetic radiation in microwave ovens were partly carburized in order to limit, during the time of operation, the volatilization of thorium. The aim of this operation is to create a layer of carbide at the surface of the filament, using a benzene source. The carburization of tungsten being a reversible chemical reaction at high temperature, it is quite possible to carry out carburization-decarburization cycles. The interest of this work is to study some parameters leading to the decarburization of partly carburized filaments in order to avoid this phenomenon during fabrication and normal use of magnetrons.

The effects of pulsed laser annealing of CdTe (containing small quantities of Zn or Hg) are described for single crystals and thin film samples. The influence of laser power, wavelength, number of applied pulses, and homogenization on the morphological changes and some physical parameters are investigated with the help of scanning electron microscopy (SEM), x-ray diffraction (XRD), UV/VIS absorption spectrophotometry and (photo)electrochemical characterization. Melting of the sample surface was observed in most cases at the laser pulse energies used; the surface modifications as seen by SEM depended in a characteristic way on laser wavelength and photon density. Use of homogenized beams led to a restructuring of the surface, entailing in-depth melting with the formation of holes and spongy structures, in particular upon multi-pulse irradiation at lambda =532 nm. In general, surface melting promoted the healing of cracks which were sometimes formed by laser-induced thermal shock in the crystal plates (such cracks were not found in the thin films). Surface melting and material loss by evaporation is accompanied by some amorphization and maybe caused by a change in doping profile and/or stoichiometry near the surface.

This paper reports on the density of silica aerogels after low- and high-temperature supercritical drying as well as on the corresponding optical refractive index. The influence of adsorbed water or methanol has also been investigated. Aerogel samples are prepared via sol-gel process of pH 11 and pH 13 tetramethoxysilane/methanol/water solutions using ammonia as catalyst. After high-temperature drying, the pH 11 aerogels show volume shrinkage of less than 10%, while pH 13 aerogels and aerogels dried at low temperature with carbon dioxide shrink by about 50%. A linear relationship between refractive index and density is observed. Its slope is (0.221+or-0.002)*10^-3 m³ kg^-1 and (0.193+or-0.002)*10^-3 m³ kg^-1 before and after thermal treatment at 500 degrees C for 14 h in air, respectively. The larger value is due to hydroxyl or methoxy groups in the aerogel. Adsorbed water or methanol cause the index to increase, as do molecules. The distribution of SiO₂ on the nanoscopic level does not influence the index.

In this paper we study the problem of electromagnetic fields around a conductor bar of rectangular cross section. A thin-skin model is established for calculating such fields. It is revealed that the behaviour of fields near the edges of the bar is dependent on a material property and frequency-related dimensionless parameter m. The current distribution on the surface of the conductor tends to be uniform if m is small and edge-concentrated if m is large. Experiments for testing the theory are carried out. The results of these experiments are also presented, which are found to be in good agreement with the theory in general.

A cobalt-neodymium liquid alloy ion source for applications in focused ion beam devices was developed and investigated. The performance and the main parameters of the ion source are presented.